0:00:15good afternoon that would be present in the were carried out that the at the
0:00:19university of mister
0:00:21and i present in the age related voice disguise anything about speaker verification and as
0:00:27we know when we ask that in the performance of a
0:00:30of automatic speaker verification system we want to assess
0:00:35inter speaker variability and we wanted to more
0:00:38with h really that these guys and for that we define it that's the intentional
0:00:43modification of the speaker voice sounded like the younger one and all the first and
0:00:47so forth these we believe that the corpus with six stick speakers all native finnish
0:00:52in that's to the environment and the speech is read text in finnish and in
0:00:57and for these we got twenty six segments on each voiced i natural voice for
0:01:02voicing your voice into different sessions for the speakers at the same time we do
0:01:08the prior to recording with that too smart phones just to check this channel differences
0:01:15and so we for the performance evaluation we use just a
0:01:19mfcc feature fifty four dimensional and to a s p system base gmm ubm and
0:01:26i-vector system with cosine and b lda score and ported that's there are we used
0:01:31to going for the three conditions basically because we have one not recognition and then
0:01:35the disguised condition would be different all and
0:01:38own voice
0:01:39we got this number of trials for each of these conditions and the channels as
0:01:44i already mentioned so they training they that all these microphone data from the natural
0:01:51voice and the corresponding disguised voice was used for testing in the in the disguise
0:01:59so this is nick a big of the results and we can see for the
0:02:03i-vector be lda a system we can see i degradation between the natural voice baseline
0:02:10of the system and then they these guys all and these guys don't we can
0:02:14see this degradation in for more details on the experimental setup and
0:02:19the data please this poster seven